This application provides continuing data management support and statistical support for the Consortium to Establish a Registry for Alzheimer's Disease (CERAD). Data base support is required for the CERAD data collection, one of the largest active longitudinal data bases of Alzheimer's disease (AD), with more than 1,200 cases and 470 healthy elderly controls. This large data base contains annual demographic, clinical, neurological, and behavioral data, and, ultimately, neuropathological data, for probable and possible AD cases enrolled since 1987. This case material is being expanded as annual follow-up and autopsy data accumulates. Statistical support will encompass four general areas: 1) analyzing, interpreting, and publishing the wealth of baseline and longitudinal information accumulating in the CERAD data base; 2) development, description, and application of statistical methods for analyzing data on AD; longitudinal analyses and survival analytic methods will be a special concern; 3) extraction and reporting of information from CERAD relevant to the design and analysis of clinical trials in AD; power calculation and sample size requirements based on different proposed study designs will be estimated and compared using the CERAD data base.